1. Field of the Invention
The present invention relates to an apparatus for detecting a sign of failure of a vehicle-mounted device.
2. Description of Related Art
It is known to make a diagnosis to detect failure of a vehicle-mounted device, and record the result of the diagnosis in a nonvolatile memory. To make such a diagnosis, it is known to make a determination that failure has occurred only after an abnormal state was detected more than once in order to prevent an erroneous determination. For example, refer to Japanese patent application Laid-open No. 2004-330892 (patent document 1). In this technique, the number of times of occurrence of an abnormal state is recorded at system off. Here, “an abnormal state” is a state in which the value of a parameter of a device is outside a normal range, and “make a determination of occurrence of failure” means to record occurrence of failure as diagnostic data after an abnormal state has continued longer than a predetermined period of time, or occurred more than a predetermined number of times.
It is also known to detect an abnormal state of an oxygen sensor based on a magnitude of the area value on the rich/lean side of its sensor value. For example, refer to Japanese patent application Laid-open No. H9-137717 (patent document 2).
Incidentally, there is a demand to enable a user of a vehicle or a mechanic to estimate a time to make a repair or replacement for each of vehicle-mounted devices, in order to assure safe driving of the vehicle and prevent reduction of fuel economy by repairing or replacing them before they fail.
In the technique disclosed in patent document 1, the number of times of occurrence of an abnormal state is counted to determine whether a device has failed or not. However, there may occur a case where the device is erroneously determined to have failed before it fails actually, for example, if the value of a parameter of the device oscillates between a normal range and an abnormal range, and accordingly the abnormal state is detected a number of times during a short period of time.
On the other hand, in the technique disclosed in patent document 2, it is possible to determine that a device is on the brink of failure based on the detected area value. However, since the information that the device is about to fail is not recorded and lost when the detection system is powered off, it is not possible to use this information in any control using the output of the oxygen sensor, or in the next fault detection.
As explained above, the technique disclosed in patent document 1 has a problem in that a device operating normally may be replaced for no reason, while the technique disclosed in patent document 2 has a problem in that a device on the brink of failure may not be repaired or replaced at an appropriate timing.